Environ. Sci. Technol. 2001, 35, 3691-3694
Time-Resolved Laser Fluorescence Spectroscopy Study of the Sorption of Cm(III) onto Smectite and Kaolinite THORSTEN STUMPF* Forschungszentrum Rossendorf, Institut fu ¨ r Radiochemie, P.O. Box 510119, D-01324 Dresden, Germany ANDREAS BAUER Forschungszentrum Karlsruhe, Institut fu ¨ r Nukleare Entsorgung, P.O. Box 3640, 76021 Karlsruhe, Germany FREDERIC COPPIN University Paul Sabatier, UMR 5563 - LMTG, 38 rue des 36-Ponts, 31400 Toulouse, France JAE IL KIM Forschungszentrum Karlsruhe, Institut fu ¨ r Nukleare Entsorgung, P.O. Box 3640, 76021 Karlsruhe, Germany
For long-term performance assessment of nuclear waste repositories knowledge concerning interactions of actinides with mineral surfaces is imperative. The mobility and bioavailability of released radionuclides is strongly dependent on sorption/desorption processes onto mineral surfaces. Therefore it is necessary to characterize the surface species formed and to elucidate the reaction mechanisms involved. The high fluorescence spectroscopic sensitivity of Cm(III) has attracted our interest regarding the complexation process of Cm(III) onto smectite and kaolinite as a model system for the sorption of trivalent actinides in the trace concentration range. We conclude that at low pH Cm(III) is sorbed onto kaolinite and smectite as an outer-sphere complex and retains its complete primary hydration sphere. With increasing pH inner-sphere adsorption onto kaolinite and smectite occurs via the aluminol edge sites. The same evolution of the Cm(III)-clay surface species as a function of pH was observed for both minerals. Starting at a pH g 5 we observe the formation of a tAl-O-Cm2+(H2O)5 surface complex, which is replaced by a second species at higher pH. The second surface complex may be a monodentate tAl-O-Cm+(OH)(H2O)4 species or bidentate t(Al-O)2-Cm+(H2O)5 species. The Cm(III)/clay surface complexes are characterized by their emission spectra (peak maxima at 598.8 and 603.3 nm) and their fluorescence lifetime (both 110 ( 7 µs). An important result in view of the mobility and bioavailability of radionuclides is that no incorporation of Cm(III) into the bulk clay structure was observed.
Introduction Sorption to mineral surfaces is a major process controlling the concentration, mobility, and bioavailability of radionu* Corresponding author phone: (0049)351-260 2432; fax: (0049)351-260 3553; e-mail:
[email protected]. 10.1021/es001995o CCC: $20.00 Published on Web 08/16/2001
2001 American Chemical Society
clides in nature. Oxide and clay minerals are the main constituents in groundwater and soil systems responsible for metal ion sorption. In most cases, experimental sorption data have been macroscopically interpreted through empirical means, such as partition coefficients and adsorption isotherms (e. g., ref 1 and references therein). The results from surface sorption models are increasingly employed to describe the sorption process. Despite their usefulness in appraising sorption mechanisms their application has been applied mostly to simple oxide or hydroxide minerals. This mainly results from the lack of information on the chemical composition of mineral surfaces and mineral water interfaces (2). As clay minerals have a complex surface structure, they show complicated sorption characteristics, especially at low ionic strength (3). An application of surface complexation models (SCM) for clays is not as successful as for oxides or hydroxides. Therefore molecular level spectroscopic information is needed to identify the sorption mechanisms including the structure of the species involved in the sorption process and to validate the applicability of SCM. Kaolinite is a 1:1 layer type clay mineral in which each Si-, Al-hydro(oxide) component exists in two distinct structural environments at the surface. There is only minimal substitution of variable valence cations in the structure and hence only minor permanent structural charge. In contrast to kaolinite in smectite each aluminous octahedral layer is bound to two silica rich tetrahedral layers. Smectite is an expandable 2:1 layer type clay mineral that has both interlayer sites and ionizable hydroxyl sites on its external surface for the metal cation sorption. Substitution of Mg2+ and Fe2+ atoms for Al3+ in the octahedral layer creates a positive charge deficit giving the overall structure a net negative charge. The O and OH atoms at the edges of kaolinite and smectite, or at the gibbsite basal plane of kaolinite, become hydrolyzed and form Lewis base or Lewis acid functional groups, which are the source of the pH dependent charge (4). Binding to these hydroxyl sites on the external surface is thought to be a complexation reaction similar to that occurring at the oxide mineral surface. The aim of this study is to verify this assumption and to investigate the sorption mechanism of metal cations onto kaolinite and smectite. As representative for trivalent actinides curium was chosen, because its fluorescence spectroscopic sensivity enables speciation studies at very low concentrations. The limit of detection for Cm3+(aq) by time-resolved laser fluorescence spectroscopy (TRLFS) is 5 × 10-12 mol/L (5), and thus speciation may be performed in the subnanomole concentration range. TRLFS has been proven as a versatile tool to study both solution chemistry of curium (6-11) and surface complexation of Cm(III) onto silica (12) and γ-alumina (13). A review of the spectroscopic properties of Cm(III) is given elsewhere (5). In this paper the speciation of surface sorbed metal ions was studied by TRLFS in the nanomolar concentration range, as a function of pH at low ionic strength.
Experimental Section Materials and Methods. Well crystallized, fine grained kaolinite (e 2 µm) from St. Austell (UK) was used. The smectite used was < 0.1 µm fraction of the Ceca bentonite separated by sedimentation technique. The cation exchange capacity (CEC) of the initial and the reacted products was determined according to Meier and Kahr (14) and was found to be 3.7 mequiv/100 g for kaolinite and 75 mequiv/100 g for smectite. A detailed description of the clay samples is given elsewhere (15). VOL. 35, NO. 18, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3691
Solutions were prepared from analytical grade chemicals. For the experiment Cm-248 (t1/2 ) 3.4 × 105 years) was used. The stock solution consists of 97.2% Cm-248, 2.8% Cm-246, and less than 0.01% Cm-244 in concentrated HCl. The initial curium concentration was adjusted to 3.0 × 10-7 mol/L and controlled in the starting solution by ICP-mass spectroscopy. Experimental Setup. As reaction vessels 20 mL polypropylene bottles were used. Batch experiments were performed in a glovebox under argon atmosphere at 25 °C to avoid contamination with atmospheric CO2, which can lead to the formation of Cm(III) carbonate complexes. To prevent precipitation we started at pH 3.5 and increased the pH up to nine in small steps. As background electrolyte we used NaClO4 solutions (I ) 0.025 M). Clay samples (S:L 0.25 g L-1) were equilibrated at pH 6.7 for 1 week in the dark. Kinetic tests were performed at different Eu(III) trace concentrations (Eu(III) serves as a Cm(III) homologue) as a function of pH for times between 1 and 30 days. No kinetic effects were observed, i.e., a steady state was reached within 1 day. Nevertheless, batch samples were stored for at least 2 days and shaken periodically. The pH was adjusted by adding analytical grade NaOH or HClO4. TRLFS measurements were performed using a pulsed Nd: YAG pumped dye laser system (Continuum, Powerlite 9030, ND 6000). Details on the experimental setup are given elsewhere (16-18). The system consists of a polychromator (Chromex 250) with a 1200 lines/mm grating. The emission spectrum of Cm(III) in aqueous kaolinite and smectite suspensions was recorded at room temperature in the 580620 nm range, within a constant time window of 1 ms exciting at 396.6 nm (laser dye: Exalite 398). The emission spectrum was recorded 1 µs after the laser pulse to fade out any background fluorescence. For measuring the emission decay, the delay time between laser pulse and camera gating was scanned with time intervals between 10 and 20 µs.
FIGURE 1. Fluorescence emission spectra of 3.0 × 10-7 mol/L Cm(III) in aqueous kaolinite (a) and smectite (b) suspensions at various pH; spectra are scaled to the same peak area.
Results and Discussion The fluorescence emission band of the Cm(III) aquo ion has a peak maximum at 593.8 nm at pH e 6. The evolution of the fluorescence emission spectra of Cm(III) with kaolinite in aqueous suspension at various pH is shown in Figure 1a. At pH e 4 only the signal of the Cm(III) aquo ion is detected. At pH g 5, the intensity of the 593.8 nm peak decreases and two other peaks appear with maxima at 598.8 nm (pH g 5) and 603.3 nm (pH g 7.22). Comparable evolution is observed for the smectite suspension (Figure 1b). With increasing pH, the intensity of the 593.8 nm peak decreases and the same red shifted peak maxima appear at 598.8 nm (g pH 5.40) and 603.3 nm (pH g 7.15). It is remarkable that, in case of smectite and kaolinite, the spectra of the sorbed Cm(III) species at low pH (e 5) show no difference to spectra of the Cm(III) aquo ion. This observation indicates that at pH values e 5 the sorbed Cm(III) ion retains its hydration sphere. Such a sorption process suggests outer-sphere complex formation on the interlayer sites. The red shift of the fluorescence emission of Cm(III) at higher pH (> 5) in the kaolinite and smectite suspensions is caused by a change in the ligand field of the Cm(III) ion and indicates inner-sphere complex formation. A peak deconvolution was carried out to resolve the individual species from the composite fluorescence emission spectra at pH g 5. Three different species have been identified. In addition to the Cm(III) aquo ion, two different inner-sphere Cm(III) surface complexes are formed. All measured spectra have been deconvoluted using pure component spectra with peak maxima at 598.8 and 603.3 nm. In Figure 2 Cm(III) species distribution in the presence and absence of clay minerals are shown. Previous studies confirm that in the absence of kaolinite and smectite the 3692
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 18, 2001
FIGURE 2. Species distribution of Cm(III) as a function of pH in the presence and absence of kaolinite and smectite. shift of the Cm(III) emission spectrum starts with reaching pH 5 by forming the first hydrolysis complex Cm(OH)2+. The Cm3+ hydrolysis process stops not until reaching pH 8. Even at pH 7.5 more than 15% of the Cm3+ aquo ion is detectable in a carbonate free curium solution as it was shown by Fangha¨nel and Neck (6, 19). By comparing the species distribution of the hydrolysis of curium and the sorption of Cm(III) onto kaolinite and smectite one can see that the hydrolysis of the actinide plays no role in the sorption process. When hydrolysis occurs all of the Cm(III) is already sorbed.
FIGURE 3. Fluorescence emission spectra of Cm(III) in aqueous kaolinite, smectite, γ-alumina, and silica suspensions in the pH range between 8.22 and 9.56; spectra are scaled to the same peak area. That is furthermore a proof that precipitation can be neglected. The pH dependence of the spectroscopic data decay indicates that one proton is involved in the formation of the second Cm(III) surface complex from the first innersphere complex according to the reaction
complex 1 f complex 2 + H+ This is verified by a slope analysis. The concentration ratio, log(complex 2/complex 1), as a function of pH is linear with a slope of one. There are two plausible different reactions, which would involve the release of one proton. One reaction is the hydroxo complex formation according to
tX-O-Cm2+ • nH2O f
tX-O-Cm+(OH) • (H2O)n-1 + H+
the second reaction is the transformation of a monodentate to a bidentate Cm complex according to
tX-O-Cm2+ • nH2O + tX-O-H f
(tX-O)2-Cm+ • nH2O + H+
The peak maximum at 603.3 nm in the kaolinite and smectite spectrum at pH 8.22 corresponds to the second Cm(III) surface complex, which was also found in the alumina system (13). Fluorescence emission spectra of Cm(III) in kaolinite, smectite, γ-alumina (13), and silica (12) suspensions between pH 8.22 and 9.56 are shown in Figure 3 together with the spectrum of Cm(III) aquo ion at pH 1.89. The similarity of spectra of sorbed curium measured in γ-Al2O3 and clay suspensions and their difference compare to the spectrum of sorbed Cm(III) in silica suspension suggest that innersphere complexation occurs via the aluminol sites. Fluorescence decay rates are caused by radiative and nonradiative processes. For the 6D7/2 f 8S7/2 transition of Cm(III) in 1 M HClO4, the decay rate is calculated to be 770 s-1 (20). This corresponds to a radiative lifetime (reciprocal decay rate) of 1.3 ms. Nonradiative decay is mainly due to energy transfer from the excited state to ligand vibronics, e.g., OH vibration of coordinated H2O or OH- molecules. A lifetime of 68 ( 3 µs is determined for the Cm(III) aquo ion (21-23). The increase in lifetime by complexation reflects the exclusion of water molecules of the first coordination sphere of the Cm(III). Using the method developed by Kimura and Choppin a lifetime of 68 µs, determined for the Cm(III) aquo ion, corresponds to nine water molecules, and a value of 1.3 ms corresponds to zero H2O molecules in the first coordination shell of Cm(III) (16).
FIGURE 4. Schematic representation of Cm(III) sorption onto smectite. TOT refers to tetrahedral/octahedral/tetrahedral sheets. INT is the abbreviation for the interlayer sites. (A) At low pH Cm(III) is sorbed as an outer-sphere complex onto smectite and retains its complete primary hydration sphere. (B) Starting at a pH g 5 a t Al-O-Cm2+(H2O)5 surface complex is formed. (C) The first innersphere complex is replaced by a second species at higher pH. The second surface complex may be a monodentate tAl-O-Cm+(OH)(H2O)4 species or a bidentate t(Al-O)2-Cm+(H2O)5 species. The fluorescence emission lifetime of Cm(III) in the kaolitite and smectite suspensions at various pH was determined. In both cases at pH e 5, the number of water molecules in the first coordination shell of the Cm(III) was found to be 9. This proves that the outer-sphere complex has not only the same fluorescence emission spectrum like the Cm(III) aquo ion with the peak maximum at 593.8 nm but also the same number of water molecules in the first coordination shell of curium. As the pH increases, innersphere adsorption becomes more important. The fluorescence emission lifetime increases from 68 ( 3 µs to 110 ( 7 µs corresponding to a decrease in the number of water molecules in the first coordination shell of about 5.0 ( 0.4 VOL. 35, NO. 18, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3693
at pH g 7. As a conclusion the Cm(III) species adsorbed onto smectite and kaolinite are illustrated schematically in Figure 4. The fluorescence lifetimes of Cm(III) sorbed onto kaolinite, smectite, and γ-alumina are very similar. The lifetime of 110 ( 7 µs was observed for Cm(III) sorbed onto these minerals corresponding to five water molecules in the first coordination shell of the sorbed Cm(III) and differs from the lifetime found for a Cm(III) silica suspension at pH g 8.7 (12). The long Cm(III) fluorescence lifetime observed in the silica system at high pH indicates that curium has lost its primary hydration sphere during the sorption process and is imbedded into the silica structure. By comparison, this also suggests that Cm(III) is not incorporated into the clay structure. Our results give a detailed description of the sorption mechanisms of Cm(III) onto kaolinite and smectite in the trace concentration range as a function of pH. This insight into the sorption mechanisms of a trivalent radionuclide onto a clay mineral surface will be of importance for the predictive modeling of radionuclide migration.
Literature Cited (1) Bradbury, M. H.; Baeyens, B. Geochim. Cosmochim. Acta 1999, 63, 325. (2) Hyun, S. P.; Cho, Y. H.; Kim, S. J.; Hahn, P. S. J. Colloid Interface Sci. 2000, 222, 254. (3) Bayens, B.; Bradbury, M. H. J. Contam. Hydrol. 1997, 27, 199. (4) Strawn, D. G.; Sparks, D. L. J. Colloid Interface Sci. 1999, 216, 257. (5) Kim, J. I.; Klenze, R.; Wimmer, H. Eur. J. Solid State Inorg. Chem. 1991, 28, 347.
3694
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 18, 2001
(6) Fangha¨nel, T.; Kim, J. I.; Paviet, P.; Klenze, R.; Hauser, W. Radiochim. Acta 1994, 66/67, 81. (7) Paviet, P.; Fangha¨nel, T.; Klenze, R.; Kim, J. I. Radiochim. Acta 1996, 74, 99. (8) Ko¨nnecke, Th.; Fangha¨nel, Th.; Kim, J. I. Radiochim. Acta 1997, 76, 131. (9) Fangha¨nel, T.; Weger, H. T.; Schubert, G.; Kim, J. I. Radiochim. Acta 1998, 82, 55. (10) Aas, W.; Steinle, E.; Fangha¨nel, Th.; Kim, J. I. Radiochim. Acta 1999, 84, 85. (11) Fangha¨nel, Th.; Ko¨nnecke, Th.; Weger, H.; Paviet-Hartmann, P.; Neck, V.; Kim, J. I. J. Solution Chem. 1999, 4, 447. (12) Chung, K. H.; Klenze, R.; Park, K. K.; Paviet-Hartmann, P.; Kim, J. I. Radiochim. Acta 1998, 82, 215. (13) Stumpf, Th.; Rabung, Th.; Klenze, R.; Geckeis, H.; Kim, J. I. J. Colloid Interface Sci. 2001, 238, 219. (14) Meier, L.; Kahr, G. Clays Clay Miner. 1999, 47, 386. (15) Bauer, A.; Berger, G. A. Applied Geochem. 1998, 13, 905. (16) Kimura, T.; Choppin, G. R. J. Alloys Comp. 1994, 213/214, 313. (17) Wimmer, H.; Klenze, R.; Kim, J. I. Radiochim. Acta 1992, 56, 79. (18) Klenze, R.; Kim, J. I.; Wimmer, H. Radiochim. Acta 1991, 52/53, 97. (19) Neck, V.; Fangha¨nel, Th.; Kim, J. I. Bericht FZKA 6110 1998. (20) Carnall, W. T.; Crosswhite, H. M. Report ANL-84-90 1995. (21) Beitz, J. V.; Hessler, J. P. Nucl. Techn. 1980, 51, 169. (22) Beitz, J. V.; Bowers, D. L.; Doxtader, M. M.; Maroni, V. A.; Reed, D. T. Radiochim. Acta 1988, 44-45, 87. (23) Fangha¨nel, T.; Kim, J. I.; Klenze, R.; Kato, J. J. Alloys Comp. 1995, 225, 308.
Received for review December 20, 2000. Revised manuscript received June 25, 2001. Accepted June 27, 2001. ES001995O
